Zhuzhou CRRC Times Electric Co., Ltd. (3898.HK) - Deep Dive Research Report

Prepared 2026-07-06. All figures in RMB unless stated. No valuation, price, or market-cap data for the subject company is included, per mandate.

A note on reporting cadence and the most recent period

Zhuzhou CRRC Times Electric is dual-listed: H-shares on the Hong Kong Stock Exchange (3898.HK, listed 2006) and A-shares on the Shanghai STAR Market (688187.SH, listed 2021). As a Chinese A-share issuer it files on a quarterly cadence - Q1 (end-April), interim/H1 (end-August), Q3 (end-October), and the annual report (end-March/April) - but it holds full investor briefings (the nearest thing to a Western earnings call) only twice a year, at the interim and annual results.

Working from that calendar, the most recent release should be Q1 2026 (quarter ended 31 March 2026), which was due late April. It was published on 27 April 2026 (revenue +12.45% to RMB 5.10bn, net profit +1.91% to RMB 643m). The H1 2026 interim is due around late August 2026 and has not yet been released; there is no evidence to the contrary, only the ordinary forward calendar. This report therefore anchors to Q1 2026 and works back through the prior five reporting periods.

The six reporting periods used throughout: Q1 2026 (27 Apr 2026), FY2025 annual (late Mar 2026, briefed subsequently), Q3 2025 (Oct 2025), H1 2025 interim (Aug/Sep 2025), Q1 2025 (Apr 2025), and FY2024 annual (Mar 2025, Hong Kong investor briefing 13 May 2025). A caveat that applies to Section 9: this issuer does not publish verbatim quarterly transcripts in the way a US company does. The interim and annual investor briefings are the substantive management-commentary events; the Q1 and Q3 releases are numbers-plus-summary filings. Where I quote or paraphrase management I am drawing on those briefings and results announcements, and I flag where a full transcript was not available.

Section 1: What the company does

Strip away the corporate name and this is, at its heart, the company that makes the "brains and muscles" that move electric trains. When a high-speed train, a metro car, or an electric locomotive draws power from an overhead line, that raw electricity is useless until something converts it, controls it, and delivers it to the motors at exactly the right voltage, frequency, and torque. That conversion-and-control hardware - the traction converter, the traction motor, the network control system that coordinates every car - is what Zhuzhou CRRC Times Electric builds. For most of the trains that CRRC (the state-owned parent, the largest rolling-stock manufacturer in the world) rolls out of its factories, Times Electric is the supplier of that electrical drivetrain.

The company traces to the CRRC Zhuzhou Institute (Zhuzhou, Hunan province), a research institute founded in 1959 to develop traction and electrical technology for Chinese railways. That institutional DNA matters: the company was an R&D house first and a manufacturer second, which is why it accumulated deep semiconductor and power-electronics know-how decades before that became fashionable. It was carved out and listed in Hong Kong in 2006, and added a STAR Market A-share listing in 2021.

The core value proposition is this: China's rail network is enormous and still growing, every electric train needs a traction system, and Times Electric has supplied that system to the domestic fleet for so long, across so many train models, that it has ranked first in China's urban-rail traction-converter market for 13 consecutive years (2012-2024) and holds well over half the domestic market. Trains are safety-critical, certified assets that run for 30 years; you do not swap out the traction supplier casually. That is the moat.

What makes the product genuinely hard is the vertical stack underneath it. A traction converter is built around high-power semiconductor devices - IGBTs (insulated-gate bipolar transistors) and, increasingly, silicon-carbide (SiC) devices - that switch thousands of amps at high voltage. Very few companies in the world can design and fabricate these chips and build the modules and build the converters and integrate them into a train. Times Electric can. It acquired the UK's Dynex Semiconductor (majority stake in 2008, full ownership 2019) to get the silicon expertise, and in 2014 it commissioned China's first, and the world's second, 8-inch IGBT production line. It is, by its own description, the only Chinese firm holding thyristor, diode, IGCT, IGBT, and SiC device technology plus power-assembly capability under one roof - a complete "wafer to module to converter to application" chain.

That chain is now the engine of the company's second act. The same power semiconductors that switch current on a train also switch current in an electric-car inverter, a wind-turbine converter, a solar inverter, and a ship's electric-propulsion drive. So Times Electric is pushing a "Rail Plus" strategy: keep the dominant, cash-generative rail business, and redeploy the underlying power-electronics platform into new-energy vehicles, renewables, marine, and industrial markets, with the stated ambition of reaching rough parity between rail and non-rail revenue by around 2027.

A concrete walk-through of what they actually do: CRRC's Sifang or Changchun factory wins an order to build a fleet of high-speed EMUs for China Railway. Times Electric supplies the traction converter (the box that turns 25kV AC catenary power into controlled three-phase power for the motors), the traction motors themselves, the auxiliary power supply, and the train network control system that lets the driver's throttle command propagate to every powered axle. Inside that converter sit IGBT modules that Times Electric fabricated in Zhuzhou from wafers it also made. The fleet then runs for decades, and Times Electric earns a long tail of spare-parts and overhaul revenue servicing the installed base.

Section 2: Business segments

The company reports in two segments: Rail Transit Equipment and Emerging Equipment. That two-box split understates the diversity, because "Emerging Equipment" is really a portfolio of five or six distinct businesses that happen to share the power-electronics platform. I treat each meaningful business as its own sub-section.

Rail Transit Equipment (~60-65% of revenue)

This is the historic core and still the majority of the company. It bundles the rail-transit electrical equipment (traction converters, traction motors, auxiliary converters, train network and control systems, and braking systems), rail-transit engineering machinery (track-laying and maintenance machinery), and communication-signalling businesses. In FY2024 this segment grew revenue ~22%, driven by both the electrical-equipment and communication-signal lines.

Core capability: the ability to design a complete traction chain for any rolling-stock type - metro, light rail, high-speed EMU, electric and diesel-electric locomotive - and to do so with in-house semiconductors. This took the institute six decades to build. A new entrant cannot simply hire engineers; it would need the fabrication line, the qualified device library, the model-by-model type approvals, and the track record on a safety-critical asset.

Why it stands alone: it sells to a fundamentally different customer (rolling-stock builders and railway operators) under long-cycle, tender-based procurement governed by railway certification, and its economics are tied to national rail capex rather than consumer or industrial cycles.

Competitive position: domestically dominant - the incumbent traction supplier to its own parent group and the wider Chinese fleet, first in urban-rail traction for over a decade. Internationally it competes with the traction arms of Alstom, Siemens Mobility, Hitachi Rail, and Wabtec, where it is a challenger rather than the incumbent and is often gated by "buy-local" content rules in Europe and North America.

Role in the group: the cash cow and the technology base. It funds the R&D and capex that the emerging businesses draw on, and its steady rail-capex-linked cash flow supports the dividend.

Emerging Equipment - Power Semiconductor Devices

The crown jewel of the second act, run through the CRRC Times Semiconductor operation. It fabricates IGBTs (up to 8-inch), IGCTs, thyristors, diodes, and SiC devices, and packages them into modules for rail, automotive, renewable, and industrial customers. In H1 2024 power semiconductors alone did RMB 1.75bn, up ~27% year on year - the fastest-growing line in the group. This is the business that turns a captive rail input into a merchant product sold to third parties, including EV makers and inverter builders.

Core capability: merchant high-power silicon and SiC, domestically fabricated. In a China obsessed with semiconductor self-sufficiency, a homegrown 8-inch IGBT line and a working SiC device programme are strategically valuable and hard to replicate.

Competitive position: domestically it is one of the two or three credible high-power players alongside StarPower Semiconductor and BYD Semiconductor; globally it sits well behind Infineon, Mitsubishi Electric, Fuji Electric, and onsemi in the merchant IGBT/SiC market, but it is the vehicle for import substitution in China.

Emerging Equipment - New Energy Vehicle Electric Drive

Electric-drive systems (motor, inverter, controller, and increasingly integrated "multi-in-one" drive units) for passenger and commercial NEVs. This leverages the same inverter and semiconductor know-how used in traction. It is a high-growth, brutally competitive market: China's electric-drive component supply is consolidating around a handful of full-chain suppliers, and Times Electric competes with Inovance, BYD's in-house arm, UAES, and others, in a segment where pricing pressure is severe.

Emerging Equipment - New Energy (Wind, PV, Storage, Hydrogen)

Wind-power converters, PV inverters, energy-storage systems, and hydrogen power supplies. Again the same power-conversion platform, aimed at China's renewable build-out. Cyclical and tied to wind/solar installation volumes, and competitive against dedicated inverter specialists such as Sungrow and the wind-converter arms of turbine OEMs.

Emerging Equipment - Marine, Industrial and Sensors

A cluster of smaller businesses: marine engineering equipment (electric ship propulsion and onboard power - a beneficiary of maritime electrification), industrial electrification products (industrial drives and converters), and sensors (a genuine niche - CRRC-heritage sensor technology sold into industrial and automotive uses). These are the optionality bets: individually small, collectively evidence of how far the power-electronics platform can stretch.

Section 3: Products and business detail

Traction converter systems. The flagship product: the power-electronic box that converts catenary power into controlled motor drive. Times Electric produces variants spanning metros, light rail, high-speed EMUs (including the Fuxing family), and electric/diesel-electric locomotives. Certification is model-specific and railway-authority-governed; each design must pass type approval before it can enter service, which is why an incumbent library of approved designs is itself a barrier.

Traction motors and auxiliary systems. The motors that the converter drives, plus the auxiliary converters that power onboard systems (HVAC, lighting, doors).

Train network and control systems (TCMS). The digital nervous system coordinating the powered cars and relaying driver commands - safety-critical software and hardware.

Braking systems. Times Electric expanded into rail braking (historically a Faiveley/Knorr-Bremse-style specialty), broadening its share of the value of each train.

Communication and signalling. Rail communication and signalling equipment, a business that grew notably in FY2024.

Power semiconductor devices. The device catalogue: thyristors, rectifier diodes, IGCTs, IGBT chips and modules (including automotive-grade and up to 8-inch wafer processing), and SiC MOSFET devices and modules. Manufactured in Zhuzhou across a complete "wafer - module - assembly - converter - application" chain, with Dynex in the UK as an R&D and packaging node. The 2014 8-inch IGBT line was the milestone that reduced China's dependence on imported high-power chips; subsequent phases have added SiC and automotive capacity, with further expansion (often described as a "Phase III") extending 8-inch IGBT and SiC output into 2026.

NEV electric-drive units. Motor, inverter, and controller assemblies, increasingly integrated into multi-function drive units for carmakers.

New-energy converters. Wind-power converters, PV/string inverters, energy-storage PCS, and hydrogen power supplies.

Marine and industrial. Electric-propulsion drives for ships, industrial power-conversion equipment, and a sensor product line.

Geography. The overwhelming majority of revenue is domestic Chinese, matching the parent's rail volumes and China's NEV/renewable build-out. Exports of rail equipment follow CRRC's overseas project wins (Asia, the Middle East, and selectively Europe), while power semiconductors and drive systems are beginning to reach international customers. Dynex gives a UK footprint for semiconductor R&D and Western-market credibility.

Section 4: Customers

The customer base splits cleanly along the two segments, and the buying relationships could hardly be more different.

Rail customers. The dominant buyer is the CRRC group itself - Times Electric supplies traction and electrical systems into the rolling stock that CRRC's assembly subsidiaries build for China State Railway Group (China Railway) and for city metro operators. The end demand is set by national railway capex and municipal metro build-out. The decision-maker is the rolling-stock builder's procurement and engineering function, constrained by railway certification: a traction system must be type-approved for the specific train model, and the operator wants proven reliability on a 30-year asset. Sales cycles are long and tender-based. Switching costs are very high - re-qualifying a new traction supplier on a certified train platform is expensive, slow, and risky, which is precisely why Times Electric has held its share for over a decade. Concentration on the parent/China Railway is real, but it reflects the structure of a state-controlled rail industry rather than a fragile commercial dependency; it also means rail revenue rises and falls with the national railway equipment and maintenance budget.

Power-semiconductor and drive customers. Here the relationship is merchant and competitive. Buyers are EV makers, inverter and turbine builders, and industrial OEMs. The decision-maker is an engineering/sourcing team choosing on device performance, price, supply security, and - crucially in China - the appeal of a domestic alternative to Infineon. Qualification still creates stickiness (automotive-grade power devices must pass reliability qualification and be designed into a platform), but pricing is far more contested than in rail, and second-sourcing is common. This is lower-switching-cost, higher-competition business than rail.

Contract structure. Rail is milestone/tender-based project revenue with a long spare-parts-and-overhaul tail from the installed fleet, giving reasonable predictability tied to the rail budget cycle. Emerging-equipment revenue is more spot- and volume-driven and more exposed to price competition and end-market cyclicality.

Section 5: Competitive landscape

The competitive picture is best understood segment by segment, because Times Electric is a dominant incumbent in one arena and a challenger in the others.

In rail traction, domestically it is the entrenched leader - the captive and preferred supplier to the world's largest rolling-stock group, first in urban-rail traction for 13 straight years, with the certification library and installed base that make displacement impractical. Internationally it competes with the traction/propulsion arms of Alstom, Siemens Mobility, Hitachi Rail, and Wabtec. Against these it wins on cost and on the scale of its home market, and loses on "buy-local" content rules, geopolitics, and incumbency in Europe and North America.

In power semiconductors, the global merchant leaders are Infineon, Mitsubishi Electric, Fuji Electric, and onsemi; domestically its peers are StarPower Semiconductor and BYD Semiconductor. Times Electric's edge is being the only Chinese firm with the full high-power device-to-application chain and a captive rail demand base to seed volumes; its exposure is that in merchant IGBT/SiC it is a share-taker, not the technology frontier.

In NEV electric drive, it is one of many in a consolidating, price-savage market led by Inovance, BYD (in-house), and others.

Barriers to entry differ by segment. In rail they are very high - decades of certified designs, safety track record, in-house silicon, and a captive customer. In merchant semiconductors and EV drives they are moderate and eroding, because capital and talent are flooding into Chinese power electronics; margins there reflect that.

Market caps are approximate peer-size references only, in each company's local currency, as of roughly July 2026, and move continuously. Several competitors (Siemens, Hitachi, Mitsubishi) are diversified groups where rail or power semis is one division, so group market cap overstates the directly-competing business.

Section 6: Industry

Rail demand drivers. The rail-equipment business rides China's railway capex - new high-speed and conventional line construction, metro build-out in dozens of cities, and the maintenance/overhaul cycle on an already-vast installed fleet. This is government-directed infrastructure spend, so it is policy-sensitive and lumpy year to year, but the maintenance tail on the world's largest high-speed network provides a growing recurring floor as the fleet ages into its overhaul years. Globally, the rolling-stock and rail-equipment market is a multi-tens-of-billions-of-dollars annual market led by CRRC, Alstom, Siemens, Hitachi, Wabtec, and Stadler.

Emerging-equipment demand drivers. These businesses are levered to two of the largest industrial trends in China: NEV adoption (China accounted for around 61% of global NEV sales in early 2026) and the renewable-energy build-out (wind, solar, storage). Both consume enormous quantities of power semiconductors and power-conversion equipment. Underlying all of it is China's drive for semiconductor self-sufficiency, which specifically favours domestic high-power IGBT/SiC suppliers as import substitutes for Infineon and the Japanese majors.

Supply-chain position. Times Electric is unusually vertically integrated - it sits from the wafer fab all the way to the finished application, which is rare and gives it supply security and cost control that module-only or converter-only rivals lack. In rail it is a systems integrator to the OEM; in semiconductors it is both a captive supplier to itself and a merchant supplier to the market.

Regulation and cyclicality. Rail is regulated by railway certification and driven by state capex - defensive but policy-dependent. The emerging businesses are cyclical (NEV demand growth is decelerating from its peak; renewables installation volumes swing with subsidy and grid conditions) and exposed to fierce price competition. The blended industry picture is therefore a defensive, policy-linked rail core plus a faster-growing but more cyclical and competitive new-energy overlay.

Tailwinds: metro expansion, an aging fleet feeding overhaul demand, NEV/renewable volume growth, and the domestic-substitution premium for Chinese power semiconductors. Headwinds: the maturity of China's high-speed network (fewer greenfield megaprojects), NEV growth deceleration, savage pricing in EV components, and Western market access limits on Chinese rail exports.

Section 7: Growth triggers

Drawn from the six most recent reporting periods and investor briefings. As noted, this issuer's substantive management commentary comes at the interim (Aug/Sep) and annual (Mar, HK briefing May) events; Q1/Q3 filings are lighter. Sources are cited to the relevant release.

• "Rail Plus" pivot toward parity between rail and non-rail (NEV/renewables) revenue by ~2027. Repeated across the H1 2025 interim and FY2024/FY2025 briefings - a multi-period, consistent strategic frame rather than a one-off. (H1 2025 interim, Aug/Sep 2025; reiterated FY2025 annual briefing, 2026.)

Management has framed the company as "shifting from a rail-centric model to Rail Plus," building "a recession-resistant portfolio" spanning NEV, renewable energy, and aerospace/marine markets. (H1 2025 interim briefing, 2025.)

• Power-semiconductor capacity expansion (8-inch IGBT + SiC), extending into 2026. The completion of further-phase capacity is intended to lift 8-inch IGBT and SiC output to feed automotive and renewable-inverter demand. (FY2024 annual briefing, 13 May 2025; referenced again in FY2025 results.)

• Power semiconductors as the fastest-growing merchant line. Management has repeatedly highlighted power-semiconductor revenue outpacing the group, with third-party (non-rail) device sales into EV and renewable customers as the key ramp. (H1 2025 interim, 2025; FY2025 annual, 2026.)

• Continued expansion of domestic and international rail market share. Management cites ongoing share gains in both domestic and overseas rail transit as new EMU, metro, and locomotive orders are delivered. (FY2025 preliminary/annual results, 2026.)

• New-energy (wind/PV/storage/hydrogen) and marine electrification as new revenue legs. These are named as growth markets the power-conversion platform is being pushed into, with marine and commercial-vehicle drive specifically flagged as Q1 2026 growth contributors. (Q1 2026 results, 27 Apr 2026 - "rail transit, automotive, and marine segments drove growth.")

• NEV electric-drive volume ramp. Continued build-out of the electric-drive business as a scale play on China NEV volumes, using shared inverter/semiconductor technology. (H1 2025 interim, 2025.)

Section 8: Key risks

Concentration on state-directed rail capex. The majority of revenue depends on China's railway equipment and maintenance budget, funnelled through the CRRC group. If national rail capex slows - and China's high-speed network is maturing, with fewer greenfield megaprojects than a decade ago - the rail segment's growth flattens. The mechanism is direct: fewer new train orders means fewer traction systems sold. The offset is the growing overhaul/spares tail on the installed fleet, but that is a floor, not a growth engine.

Margin compression in the emerging businesses. The NEV electric-drive and new-energy converter markets are ferociously price-competitive and consolidating. Times Electric is a challenger, not the leader, in both. The Q1 2026 result already showed the tension: revenue grew 12.5% but net profit only 1.9%, and operating cash flow turned negative on higher payments for goods - a signal that growth is being bought at the cost of margin and working capital. If the emerging mix keeps rising while its margins stay thin, group profitability can lag revenue.

Q1 2026 (27 Apr 2026): revenue +12.45%, net profit +1.91%, "operating cash flow turned negative due to higher payments for goods." The divergence between top-line growth and bottom-line/cash conversion is the risk made visible.

Power-semiconductor competition and technology race. The semiconductor jewel is strategically valuable but faces both domestic rivals (StarPower, BYD Semiconductor) racing to add IGBT/SiC capacity and global leaders (Infineon, Mitsubishi, Fuji) at the technology frontier. A domestic capacity glut in IGBT/SiC - a real possibility given how much Chinese capital is chasing power semis - would erode the pricing that makes this the growth engine.

Geopolitics and export access. Rail exports to Europe and North America are gated by buy-local content rules and security scrutiny of Chinese infrastructure suppliers; the Dynex UK asset and any Western semiconductor ambitions are exposed to tech-transfer sensitivity. This caps the international optionality that the Rail-Plus story implicitly relies on.

Parent-group dependency and governance. Being a captive supplier to CRRC is a source of stability and of concentration risk simultaneously; related-party dynamics and state ownership mean minority H-share holders' interests are not always the first priority, though the recent buyback-and-cancel programme is a governance positive.

Section 9: Walk the talk

Six reporting periods anchor this assessment: FY2024 annual (briefed 13 May 2025), Q1 2025 (Apr 2025), H1 2025 interim (Aug/Sep 2025), Q3 2025 (Oct 2025), FY2025 annual (Mar 2026), and Q1 2026 (27 Apr 2026). The most recent is within 90 days of today. A caveat: this Chinese issuer does not publish verbatim quarterly transcripts, so the credibility read is built from the interim/annual briefings plus the numbers in each release, not from line-by-line call transcripts. I flag that rather than manufacture quotes.

The through-line management has sold across all six periods is the "Rail Plus" strategy - grow the emerging (NEV, renewables, semiconductor, marine) businesses toward parity with rail by around 2027 while defending the dominant rail franchise. On the substance of delivery, the record is broadly consistent with the talk.

Starting at FY2024 (revenue +13.4% to RMB 24.9bn, net profit +21.8%), management guided that the emerging equipment business - and power semiconductors specifically - would be the growth driver, and pointed to capacity expansion (8-inch IGBT, SiC) as the enabler. Power semiconductors did grow ~27% in the associated first-half data, validating that specific claim. The rail segment also delivered the ~22% growth management pointed to, so FY2024 is a period where the guidance and the outcome line up.

Moving through Q1 2025 and H1 2025 (interim revenue +18% to RMB 12.21bn, net profit attributable +12.9% to RMB 1.67bn, total profit +24.1%), the company kept revenue compounding at a healthy clip and reiterated the Rail-Plus frame and the 2027-parity ambition. It also raised the interim dividend (proposing RMB 4.40 per 10 shares), a tangible follow-through on the "return capital" posture. The interim delivered the double-digit revenue growth that had been the running promise.

By FY2025 (preliminary revenue +15.5% to RMB 28.76bn, net profit attributable +10.9% to RMB 4.11bn, and notably net profit after non-recurring items +21.2%), the top-line kept its promised double-digit pace and the "quality of earnings" improved (the after-exceptionals number outgrew the headline). Management "reinforced its strategic vision" - i.e., stuck to the same Rail-Plus story it had told for two years, which is itself a credibility marker: no strategic U-turns, no quietly dropped targets.

The one place where the talk and the reality visibly diverge is profit conversion in Q1 2026: revenue +12.5% but net profit only +1.9%, with operating cash flow turning negative. Management's growth narrative held on the top line, but the quarter exposed that the emerging-business mix carries thinner margins and heavier working-capital needs than the rail core. This is not a broken promise - management never promised margin expansion - but it is the first period where the "growth" story and the "profitable growth" implication came apart, and it is the thing to watch.

Assessment: management does broadly what it says. The strategy has been stable and repeated across six periods with no reversals, revenue growth has consistently landed in the promised double-digit range, and the capital-return posture has been backed with real dividend increases and share cancellations. The honest qualifier is that the company has been more reliable on revenue delivery than on margin/cash conversion, and Q1 2026 flagged that gap. This reads as a credible, execution-focused management team that under-discusses the margin cost of its growth pivot rather than one that overpromises outright.

Section 10: Shareholder friendliness index

Dividends. The dividend has grown strongly and consistently over the last three financial years. FY2023 DPS was approximately RMB 0.55 per share. FY2024 rose to RMB 0.78 per share (RMB 7.80 per 10 shares), up ~41.8% year on year. FY2025 rose again, with an interim of RMB 4.40 per 10 shares (RMB 0.44) plus a proposed final of RMB 0.68 per share, for a full-year total of roughly RMB 1.12 - up on the order of ~44% again (2025 final dividend approval was scheduled for 26 June 2026, ex-date 30 June, payment 7 August). Three consecutive years of ~40%+ dividend growth off a rising earnings base is a clear "returns capital" signal, and the payout remains well covered by earnings rather than being propped up above 100% of profit.

Buybacks and dilution. The company has moved from paying dividends to actively retiring shares, which is the stronger form of capital return. Under a mandate approved at the 27 June 2025 AGM, it was authorized to repurchase up to 48,904,090 H shares (~3.6% of issued capital). It began repurchasing on 10 February 2026 and, between then and 6 May 2026, spent about HK$364.7m and then cancelled 10,245,600 repurchased H shares, cutting issued share capital to 1,347,702,812 shares - a genuine reduction, not treasury warehousing. Per MoatMap's disclosure feed, the buying continued through the trailing ~90 days (last MoatMap scrape 2026-07-03; note this feed is flagged stale, ~49h old at report time): roughly 11.94m H shares repurchased across 1 June - 25 June 2026 at HK$37.36-40.00, with cumulative programme progress rising past ~5.5% by 25 June 2026. The company was also seeking a fresh general H-share repurchase mandate at a 26 June 2026 class meeting. Net effect over the period: shares outstanding are shrinking, not diluting - the opposite of the option-driven creep common at growth companies.

Verdict: Returns Capital - three years of rapid dividend growth plus an actively executed buy-and-cancel programme that is reducing the share count.

Section 11: Insider activities

Hong Kong's disclosure portal (HKEX DI) is gated to automated retrieval, so per the sourcing rules this section relies on the MoatMap cross-market disclosure database as the canonical source for recent insider dealing. That feed is flagged stale (last scrape 2026-07-03 23:38 UTC, ~49h before this report), so filings from the last two days may be missing. The window covered is the last 12 months of substantial-shareholder movements.

The activity is entirely institutional substantial-shareholder dealing by two custodian/asset-manager names - JPMorgan Chase & Co. and BlackRock, Inc. - not by company directors or officers. This is an important distinction: these are index/asset-management position changes crossing HKEX disclosure thresholds, not management conviction signals. There were no director or officer open-market purchases or sales in the data.

Buys - read the signal: there were no open-market insider buys and no director/officer purchases in the window. Nothing to flag as a bullish conviction signal here.

Sells - work out the why: the two clean open-market sells are both JPMorgan (748,841 shares on 2026-06-23; 274,701 shares on 2026-05-26), each a fraction of a percent of shares outstanding. These are custodian/asset-manager portfolio adjustments, not insider signals; reason not disclosed beyond ordinary position management, and the sizes are immaterial to the float. The large volume of BlackRock "Other" entries are deemed-interest and holding-level recalculations that cross HKEX DI reporting bands - characteristic of passive-fund rebalancing rather than directional conviction.

Net assessment: the 12-month insider picture is institutional churn, not management signal. Aggregate direction is mildly net selling (two JPMorgan open-market sells, no buys), but it is concentrated in two global asset managers doing routine position adjustments, with zero director or officer transactions either way. This is best read as neutral for signalling purposes - there is no management buying to lean on and no management selling to worry about. The genuine insider-level capital signal for this company sits in Section 10 (the company itself buying back and cancelling stock), not in this shareholder-disclosure churn.

Section 12: Scenarios

Bull case. The "Rail Plus" pivot works as advertised. The 8-inch IGBT and SiC capacity comes online on schedule and the power-semiconductor business keeps compounding at the ~25-30% pace it has shown, winning design-ins at Chinese EV makers and renewable-inverter builders who want a domestic alternative to Infineon. The NEV electric-drive and new-energy converter lines scale into profitability as volumes cover their fixed costs, and marine electrification adds a third leg. Meanwhile the rail core keeps throwing off cash - the aging high-speed fleet enters its heavy-overhaul years, metro build-out continues, and export wins ride CRRC's overseas projects. By the late 2020s the company looks like a diversified power-electronics group with a defensive rail cash engine funding two or three genuine growth businesses, non-rail revenue approaches parity with rail, and the buy-and-cancel programme has meaningfully shrunk the share count. The semiconductor arm's strategic value in a self-sufficiency-obsessed China becomes increasingly visible.

Base case. Management delivers roughly what it has guided. Revenue keeps growing at a low-double-digit pace, led by emerging equipment while rail grows slowly off a high base. Power semiconductors remain the standout line but face rising domestic competition that caps pricing. The margin tension visible in Q1 2026 persists: profit grows more slowly than revenue as the mix shifts toward thinner-margin new-energy businesses, and cash conversion is lumpy. The dividend keeps rising, though perhaps not at the ~40% pace of the last three years, and the buyback continues to offset any dilution. The company ends the period recognisably itself - a rail champion with a credible, growing, but not yet fully margin-proven second act - and Rail-Plus parity slips a little past the 2027 aspiration rather than arriving on time.

Bear case. Two things go wrong at once. China's rail capex slows harder than expected as the high-speed network matures, softening the cash-cow core just as the overhaul tail proves less lucrative than hoped. Simultaneously, the emerging businesses run into a domestic power-semiconductor and EV-component glut: too much Chinese capacity chasing IGBT/SiC and electric drives collapses pricing, so the growth engine grows revenue but destroys margin, and the negative operating cash flow seen in Q1 2026 becomes a pattern rather than a blip. Export optionality stays capped by Western content rules and security scrutiny. The stock's second-act narrative deflates into "a rail supplier with a low-margin components business," dividend growth stalls, and the parent-group dependency starts to look like a ceiling rather than a floor. In this world the company is still solvent and cash-generative, but the re-rating premise - a power-electronics platform escaping its rail cyclicality - fails to arrive.